Virtual reality pubs with AI patrons were once the stuff of science fiction. But advances in AI and VR have enabled Prof Paul Best to create one at Queen’s University Belfast – and use it to treat people who experience social anxiety.
Within this lively digital space, built by a VR production company Propeer Solutions, people who usually prepare extensively for conversations can experience having an unstructured one with an AI character. “Small talk is extremely daunting for somebody with social phobia,” says Best, director of the university’s Technological Innovation in Mental Health and Education (TIME) Centre. “This virtual environment provides people with social anxiety an opportunity to practise, make ‘mistakes’ in a safe space, and build up their confidence.”
A VR pub is just one example of how academics at Queen’s are using new technologies and cross-disciplinary approaches to tackle complex human issues. Lucy Baxter, an award-winning film-maker and lecturer in film practice, is exploring how embodied VR can be used for empathy training in health, social care and justice sector settings.
Her Mental Abuse Matters VR project stems from an interest in discovering “the extent to which you could inhabit someone else’s body and mind, and the effect that might have on understanding and empathy for nebulous or emotionally complex subjects”.
Through a combination of live-action film and VR, Mental Abuse Matters lets staff experience emotional abuse through the eyes of a victim, with immersive sound helping to replicate emotions such as panic, dread and shame. Queen’s has provided funding to commercialise her research, and several public sector agencies have shown enthusiasm for using it for empathy training.
In addition to this, Baxter is currently working on a project called The Afters – a series of two films about sexual consent aimed at relationship and sexuality education. “The Afters explores the subtleties and grey areas of sexual interactions including peer and online influence, boundaries and body image, lack of communication and how inner voice affects our experience and the development of relationships for young people” says Baxter.
Treating trauma with VR
As well as his role as director of the TIME Centre, Best is a practising cognitive behavioural therapist. His diverse professional background, together with a transdisciplinary approach to research, helped him see how new technologies such as VR and AI could provide added value in a therapeutic context, “particularly given the role that both imagery and exposure have within a lot of the standard CBT treatment protocols”, he says.
Along with exploring how VR can help to treat anxiety disorders, he co-designed unique VR and 360-degree video environments with people who have experienced trauma. These environments have supported individuals in safely revisiting and re-evaluating fragmented or distorted memories formed during periods of intense threat.
It’s essentially a hi-tech version of the imaginal exposure techniques used within trauma-focused CBT. This means therapists can integrate it into their treatment sessions with relative ease. “We’re using VR to access beliefs and thoughts that might otherwise take longer to reach,” says Best, “so it has the potential to speed up certain parts of therapy.”
This could free up sessions, ensuring more people receive the treatment they need at a time when mental health services are under pressure. VR environments can also replace physical trauma-site visits, which are sometimes used in therapy to support further reappraisal of the event. “That could potentially save half a day, meaning a therapist could see three other people,” says Best.
He is now co-leading development of Northern Ireland’s first Digital Mental Health Action Plan, which aims to embed digital therapy approaches into policy. Together with colleagues at the TIME Centre, he is also investigating how digital simulations could support the next generation of social workers as well as other health and social care professionals by providing a practice environment for emotionally charged conversations with service users.
Increasing inclusion with digital instruments
The design of traditional musical instruments often unintentionally excludes people with disabilities. Under the leadership of Prof Franziska Schroeder, the Performance Without Barriers research group at Queen’s – part of SARC: Centre for Interdisciplinary Research in Sound and Music – has worked with the charity Drake Music NI to develop digital instruments that address the needs of people with disabilities.
“We have a collaborative and participatory ethos where everything starts with the musicians,” says Schroeder. “Maybe they want to make wind or ocean sounds, or they’re looking for a new way to use their arm or head movements – we will design a musical instrument around them.”
Over the past decade, several disabled musicians have used these bespoke digital instruments in live performances. Schroeder’s research group is designing adaptive AI systems that can filter out involuntary body or facial movements, so that only intentional ones trigger musical sounds. “We’re just starting this journey,” she says, “but hopefully it will be exciting for a disabled musician to be able to train their own bespoke machine learning algorithm.”
Prof Roger Woods and researchers from the School of Electronics, Electrical Engineering and Computer Science are working on the emerging area of “edge” AI – a technology that runs AI algorithms on local devices, such as smartphones or sensors, instead of sending large amounts of data to the cloud for processing – for healthcare, telecoms, manufacturing, and other key industries.
Sending every piece of data to remote cloud servers for processing is slow and energy-intensive, limiting the benefits of many AI applications while also creating sustainability and security issues. To help solve this challenge, Woods’ team has developed design methods that enable AI solutions to run efficiently on specialised computer chips called Field-Programmable Gate Arrays (FPGAs), which can be embedded at the network edge. This allows data to be analysed and acted upon in real-time rather than being sent to the cloud.
A company spun out from the research, Analytics Engines, has collaborated with major technology companies and other research institutions to further develop and deploy AI solutions. One of the research projects, in collaboration with the Tyndall National Institute, involved adding an FPGA to a one-millimetre wide fibre-optic camera used by surgeons during operations – initially to improve image quality by processing data where it’s captured, but without a corresponding increase in heat.
“Once we did that, we started to look at machine learning models [trained to detect different medical scenarios], to help identify additional critical information for the surgeon as they’re operating,” says Woods, adding that the goal is to “assist surgeons by highlighting anomalies and unusual features for further examination, not to replace them”.
Queen’s academics will no doubt uncover further opportunities for enhancing human talent and addressing complex health and social issues with new technologies. By working across disciplines, they have already found unique ways to build empathy, improve mental health services, expand musical participation and enhance medical analysis, and show how a human-centric approach to technology can improve people’s lives.
Research to Reality: a series of roundtables featuring academics from Queen’s University Belfast
Queen’s University Belfast ranks among the top 200 universities worldwide in both the QS and THE rankings
Discover how collaboration is driving cutting edge research at Queen’s University Belfast
